Ben Goldacre (Glorious Leader of the Science Nerds, God amongst Men, &c.) used an entry on his Posterous blog to draw attention to the fact that Google, in creating an image to celebrate the 240th birthday of Richard Trevithick, managed to imply that he had designed and built a steam locomotive incapable of moving. (As opposed to what Trevithick actually designed: for the most part, as I recall, steam engines that exploded in a shower of scalding water and red-hot metallic fragments – but then, the story of engineering progress has always been about trial and error.) It was an amusing post, and I very much enjoyed the process of working out what Google had got wrong. For someone who, as a teenager, made copies of simplified technical drawings of steam engines and locomotives for fun,* I was appallingly slow to work out that they had connected the cogs in such a way that the engine would be attempting to drive the wheels of the locomotive in opposite directions.
So, Ben Goldacre’s post was (as they frequently are) light-hearted and fun, and his commenters joined in with the game; I particularly liked the suggestion that the design was deliberate, in order to make sure that the loco didn’t steam off the screen. As a small addendum to his post, Goldacre (Glorious Leader of the Science Nerds, God amongst Men, &c.) linked to an article Donald Simanek had written about a similar ‘problem’ with the design on the reverse of the £2 coin. I hope I’m not being unfair if I suggest that I enjoyed Simanek’s observations rather less.
Simanek’s page focuses on the misapplication and misrepresentation of scientific and engineering principles in wider culture. This kind of article could be light-hearted and fun, too, and some of it is – it’s hard to envisage anyone taking complaints about the ‘science’ of James Bond films in earnest – but other parts of it seem to be intended far more seriously. Some of these more serious objections seem fair enough. For example, the diagram of the eye that misrepresented the refraction of light by the lens does seem like an unfortunate lack of attention to detail, especially since it appeared in a publication (Scientific American) aimed at increasing the public understanding of science. Simanek’s complaints about the design on the £2 coin seem rather less well founded, however.
As we can see, the design of the reverse of the £2 coin consists of a number of concentric circles: the innermost circle contains a wagon wheel; the next a design of interlocking cogs; the third, a printed circuit; and the outermost a diagrammatic representation of the internet. Or, at least, that’s what they’re supposed to be; if you look at the second circle in and think ‘printed circuit’ then all I can say is we interpret visual imagery in very different ways.
Simanek’s objection to the design is very specific. He only complains about the interlocking cogs motif, and his complaint centres on the fact that there are 19 cogs, and that, since each cog will turn its neighbour in the opposite direction, any mechanism containing an odd number of gears could never turn. This objection, though valid as far as it goes, seems rather peculiar to me, and for all sorts of reasons.
For a start, as Simanek himself notes, it’s not as though this would be a useful ‘machine’, even if it did have an even number of cogs; all it would do is turn. There are, I guess, two ways of thinking about this. One would be to think “OMG, how dumb are these people, they’ve made a technical drawing of a machine that won’t do anything! No wonder British industry isn’t what it used to be!!” The other would be to think “Oh, look, there’s a picture of some interlocking cogs in a circle. Clearly this isn’t a real-world machine capable of performing any practical function but it still makes for an interesting decorative motif. I can see how the artist is drawing attention to mechanical technology’s centuries-long reliance on the toothed wheel, rather than attempting a technical drawing of a particular machine.”
I have to say, I’m really quite surprised that anyone falls into the former camp. I wouldn’t have believed, until I read Simanek’s article, that anyone could mistake the design on the reverse of a coin for a technical drawing of a real-world machine. I’d have thought there’s a fairly major clue in the fact that it appears on a coin, not a set of blueprints. I’m not an engineer, but my understanding is that it would be considered fairly unusual to circulate a design via the medium of mass-produced coinage. I would have also thought that pretty much everyone in the entire world would understand that the design on a coin is mostly decorative (though it does also serve to make forgery harder), and that decorative objects are not usually expected to have a practical function.
Simanek doesn’t see it this way. He quotes, in full, an email from someone at the Royal Mint explaining (with rather more patience than I think I would have been able to muster under the circumstances) that the design is intended to represent the development of technology through the ages symbolically rather than literally, but this explanation does not placate him:
A nice answer but one that I would call “arty farty wiffle waffle” (actually I would use other words but that is a more polite version). Would it not have been nice to represent the development of technology by applying some basic physics and engineering and seeing if the damn thing worked first? I am pretty sure that is how we managed such a technological revolution. Interesting and succinct ideas stand for little in physics if they clearly incorrect. [sic.]
I think we can infer from Professor Simanek’s disdain for ‘arty farty wiffle waffle’ that he is not a frequent attendee at his local art gallery, but this is perhaps for the best. I worry that someone who complains with such palpable anger that a cogged-wheel motif on a coin has no practical use might be rendered apoplectic by an inedible painting of fruit. Still, we should make allowances; the complexities of representational art are not universally understood, as anyone who’s seen a cat look round the side of the TV for a bird that flew off screen knows.
Leaving aside such wiffle waffle, however, I’m still confused by the specificity of Simanek’s complaint. It’s only the cogged wheels that trouble him, you see, and this seems a little peculiar to me. I’ve already suggested that the ‘printed circuit’ looks (to my eyes, anyway) nothing like a printed circuit. This may be due to my unfamiliarity with such objects, but even so I’m prepared to go out on a limb and suggest that it’s highly unlikely the diagram that appears on the coin represents a functional circuit. I can’t help wonder why Simanek is not railing against this lack of rigour as well. Then there’s the wagon wheel. It certainly looks like an accurate representation of a wagon wheel, but I can’t help but feel that, without an axle and another wheel to balance it, it’s not capable of doing much except falling over; why the silence on this obvious manifestation of arty farty incompetence, I wonder? I have to say, though, both of these objections pale into insignificance compared to those that can be raised against the design motif that’s intended to represent the internet.
The internet is presented as a chain linking a number of nodes together. Some nodes can route data around the network in four directions (forwards or backwards along two pathways), while others can route it around the network in two directions (forwards or backwards along a single pathway). This is a really, really bad diagrammatic representation of the way the net works. To understand why, let’s think about what would happen in the event of node failure.
The four-way nodes are clearly more critical than the two-way ones, so what would happen if one of them were to fail? Well, as I see it, the network would experience degraded performance. It wouldn’t fail – nodes adjacent to the failed node could route data the other way round the chain – but data would have to pass over much longer cable runs, which would impact on the speed and responsiveness of the network. What about if two of the four-way nodes failed? Well, in those circumstances it seems to me that the network would fail; there would be no valid pathways for routing data between some operational nodes. In the real-world internet these nodes are server clusters, so this network design would mean that a failure of two server clusters would be sufficient to bring the internet down.
This isn’t what actually happens, of course. The internet, derived as it is from a military network, has multiple redundancy built in as a design feature. Individual nodes don’t connect only to nodes adjacent to them in a chain, but establish numerous pathways to nodes scattered throughout the network; this means that the network can withstand multiple node failures and still remain operational. Access to data hosted at failed nodes will be lost (though, in reality, much internet content is duplicated at different locations, precisely to get round this problem; separate copies of this blog are hosted in Chicago, Dallas and San Antonio, for example), but each operational node will still be able to exchange data with any other operational node. (That’s assuming, of course, the cables linking them aren’t cut; the network establishes multiply redundant pathways between nodes, but many of them run over the same physical cables, an obvious problem with the resiliency of the network, as people in Georgia and Armenia recently discovered.)
It strikes me that this is an even more fundamental ‘failure’ than the one with the cogged wheels. In that case, the complaint is that an abstract representation of a particular type of technology failed to live up to real-world engineering standards. In this case, though, the design isn’t just supposed to be an abstract representation of a type of technology (digital computer networking) but a diagram of an actual, named network. Even as an abstract representation the design ‘fails’, because it shows the chain broken at two points, once to make way for the date, and once to allow the words ‘TWO POUNDS’ to appear; the breaks make the network every bit as non-functional as the ‘machine’ with the odd number of cogs in that section of the design. The ‘offence’ is surely compounded in this case by the fact that, even if the ‘problems’ in the execution of the diagram are set aside, the network shown here is nothing like the internet it purports to represent.
Now, personally, I wouldn’t raise either objection about the internet section of the design. I think, given that it has to appear in a narrow circular band around a coin, and that it has to allow room for the date and denomination, it’s not a bad stylised representation of the internet. Unlike Professor Simanek, I also don’t have a problem regarding the coin design as fundamentally decorative, rather than useful. I’m prepared to accept that a decorative motif loosely based on a particular technology is not going to adhere to the rigorous standards of technical drawing; in fact, it strikes me as positively peculiar that anyone would think that it should. But then, beyond enjoying a slight snigger, I wouldn’t raise an objection to the odd number of cogs either.
What I’m struggling to understand is why Professor Simanek – who clearly does expect coins to function as technical drawings, and is unwilling to allow for the possibility of an object having a purely decorative function – targets his complaint so narrowly. It can’t be that he didn’t notice the problems with the schematic diagram of the internet – that’s surely unthinkable in a man so alert to mistakes and blunders, and so keen to tell the world how clever he is to have noticed them. I wonder: might it be that, at some dim, dark level, even the pedantic, humourless Simaneks of this world realise that complaints like this are more likely to put people off science and technology than turn them on to it?
* – What? Steam engines are cool. Plus, it took me a long time to discover boys. Or, at least, to work out that there were things you could do with them besides talking about steam engines…